The reason for the "copy" implementation is explained in the Documentation page describing the "largest atomic size: floating-point" for your model's Model Configuration parameters. If the data going into the rate transition block has type "double" and "double" is listed as the "largest atomic size", Simulink recognizes that the "copy implementation" is enough. If you change the "Hardware board" to something with a smaller "largest atomic size" (such as "float"), you will find that the rate transition blocks use "Buf".
Please run the below command in the MATLAB R2018b command window to get the release specific documentation on "largest atomic size: floating-point":
web(fullfile(docroot, 'simulink/gui/largest-atomic-size-floating-point.html?searchHighlight=Largest%20Atomic%20size'))
Please follow the below link to search for the required information regarding the current release:
